Characterization and Multilineage Differentiation of Domestic and Black-Footed Cat Mesenchymal Stromal/Stem Cells from Abdominal and Subcutaneous Adipose Tissue

Abstract

Transplantation of mesenchymal stem cells (MSCs) isolated from bone marrow or adipose tissue is emerging as a promising tool for cell replacement therapy and regenerative medicine in domestic and endangered animal species. Defining the differentiation capability of adipose-derived mesenchymal stromal/stem cells (AMSCs) collected from different depot sites of adipose tissue will be essential for developing strategies for cell replacement therapy. In the present study, we compared the biological characteristics of domestic cat AMSCs isolated from visceral fat of the abdominal cavity (AB) with AMSCs from subcutaneous (SQ) tissue, and the functional capability of domestic and black-footed cat (Felis nigripes) AMSCs to differentiate into other cell types. Our results showed that both domestic and black-footed cat adipose–derived stromal vascular fractions contained AMSCs. Both domestic cat AB- and SQ-AMSCs showed important clonogenic ability and the minimal MSC immunophenotype as defined by the International Society for Cellular Therapy in humans. However, domestic cat AB-AMSCs had higher percentages of cells positive for MSCs-associated cluster of differentiation (CD) markers CD90⁺ and CD105⁺ (92% and 80%, respectively) than those of SQ-AMSCs (77% and 58%, respectively). Although these results may suggest that AB-AMSCs may be more multipotent than SQ-AMSCs, both types of cells showed similar expression of pluripotent genes Oct-4 and Klf4, except for higher expression of Nanog than in AB-AMSCs, and equivalent in vitro multilineage differentiation. Under appropriate stimuli, the black-footed cat and both domestic cat AB- and SQ-AMSCs differentiated not only toward mesoderm cell lineages but also toward ectoderm cell lineage, such as neuron cell–like cells. Black-footed cat AMSCs had more capability to differentiate toward chondrocytes. These results suggest that the defined AMSC population (regardless of site of collection) could potentially be employed as a therapeutic agent for both domestic and endangered diseased or injured felids.

Introduction

Mesenchymal stromal cells (MSCs) are a heterogeneous cell population (containing multipotent cells) that can be isolated from both adult and fetal tissues and exhibit a capacity to undergo multilineage differentiation in vitro. MSCs can be found in any type of adipose tissue, and abundant numbers of adipose-derived mesenchymal stromal/stem cells (AMSCs) can be isolated from lipoaspirates (Aust et al., 2004). In humans, the isolation of AMSCs and use of their intrinsic properties have changed the traditional status of adipose tissue from that of medical waste to playing a promising role in regenerative medicine (Gimble et al., 2007). Similarly, therapeutic opportunities for using MSCs isolated from bone marrow (BM) and adipose tissue have been adopted in veterinary medicine for the treatment of injuries in domestic and endangered species (Fortier and Travis, 2011; Ribitsch et al., 2010; Volk and Theoret, 2013). Clinical transplantation of BM and AMSCs for the treatment of chronic kidney disease and chronic experimentally induced asthma in domestic cats produced a mild improvement in renal function and airway remodeling, respectively (Quimby et al., 2011, 2013; Trzil et al., 2014). Moreover, restoration of elbow function without apparent pain was observed in a trial of transplantation of adipose-derived stromal vascular fraction (Ad-SVF) into the arthritic elbow of a Florida panther at the Tallahassee Museum of Natural History (www.eatlifefoundation.net/tallymu.html).

Many rare and endangered animal species held in zoological institutions are not only predisposed to chronic joint and ligament problems but also suffer from serious diseases. For example, the black-footed cat (Felis nigripes), one of the smallest cats, suffers a high mortality rate due to kidney failure resulting from a condition known as amyloidosis (Terio et al., 2008). Traditional treatments are only palliative, managing pain and offering some comfort to the animal. Thus, MSCs transplantation is emerging as a promising tool for cell replacement therapy and regenerative medicine in rare and endangered animal species predisposed to chronic medical problems.

The domestic cat (Felis catus) is a mammalian species of particular importance from several standpoints, including their evolutionary history (Driscoll et al., 2007), as a popular companion animal, a model for human diseases (Wongsrikeao et al., 2011), and for developing assisted reproductive technologies for application to the conservation of endangered felids (Pope et al., 2006a, b). In light of the genetic similarity among felid species, domestic cats are an attractive model for the study of regenerative medicine in endangered felids. Recently, it was reported that domestic cat AMSCs isolated from subcutaneous (SQ) and epididymal adipose tissue differentiated into mesodermal cells (osteoblasts, chondrocytes, and adipocytes) when stimulated with inductive factors (Quimby et al., 2011, 2013; Webb et al., 2012; Zhang et al., 2014). In vitro–isolated AMSCs are heterogeneous cell populations despite selection via plastic adherence and successive passages.

It is known that the degree of heterogeneity can depend on depot sites of adipose tissue and that the characteristic further affects their developmental capabilities. In rabbits, the osteogenic potential of AMSCs isolated from abdominal visceral (AB) adipose tissue was reported to be greater than that of AMSC isolated from SQ adipose tissue (Peptan et al., 2006). In contrast, in humans, osteogenesis was more robust in SQ-AMSCs from the flank and thigh, as compared to that of AMSCs from AB adipose tissue (Levi et al., 2010). Therefore, it is possible that the functional capability of AMSCs to give rise to cells of multiple lineages differs between cells isolated from AB or SQ tissue. Defining the differentiation capability of AMSCs collected from different sites will be essential for developing strategies for cell replacement therapy. In the present study, we compared the biological characteristics of domestic cat AMSCs isolated from both AB or SQ tissues, and the functional capability of domestic cat and black-footed cat AMSCs to differentiate into other cell types.

Materials and Methods

Subject

Tissue donors were female domestic short-hair cats between 3 and 5 years old that were group-housed in environmentally controlled rooms at the Audubon Center for Research of Endangered Species (ACRES) and one 4-year-old male black-footed cat that was housed in an outdoor enclosure at the Freeport MacMoran Species Survival Center and recently euthanized. Fresh food was provided daily, and water was always available. All animal procedures were approved by the Institutional Animal Care and Use Committee of ACRES as required by the Health Research Extension Act of 1985 (Public Law 99-1850).

Chemicals

All chemicals were purchased from Sigma Aldrich (St Louis, MO, USA) unless otherwise stated.

Isolation and growth

Adipose tissues from black-footed cat (n = 1) and domestic cats (n = 6) were collected from the AB cavity and SQ layer by laparoscopy and biopsy, respectively, following the same protocol previously described in our laboratory (Pope, 2004), transferred directly into HEPES-buffered saline solution containing 50 μg/mL gentamicin, and stored at 4°C for <72 h before processing. Tissue samples from each animal were washed three times in Dulbecco's Phosphate-Buffered Saline (DPBS, Gibco, Grand Island, NY, USA), finely cut into ∼0.5-mm² pieces using a sterile scalpel, and incubated with 1 mg/mL collagenase II (2 mL of collagenase per gram of tissue) in a 38°C water bath shaking at 150 rpm for 20–40 min or until a homogeneous mixture was obtained. After collagenase digestion, two volumes of Dulbecco's modified Eagle medium [nutrient mixture-F12 (DMEM-F12, Gibco) supplemented with 10% (vol/vol) fetal bovine serum (FBS; Gibco), 25 μg/mL amphotericin B, 50 μg/mL gentamicin, and 100 μg/mL penicillin-streptomycin) was added to digested tissue samples.

Cell mixtures were centrifuged at 1500 rpm for 10 min, the supernatant was discarded, and the pellet containing the Ad-SVF was resuspended in 1 mL of 160 mM ammonium chloride solution to remove residual erythrocytes. The Ad-SVF mixture was immediately passed through a sterile 70-μm cell strainer (BD Bioscience, San Jose, CA, USA), diluted in 5 mL of DMEM-F12 medium, and centrifuged at 1500 rpm for 5 min. The Ad-SVF pellet was resuspended and plated in 35-mm tissue culture dish or 75-cm² tissue culture flask (Nunc, Denmark) containing 3 or 7 mL of DMEM-F12 medium supplemented with 10% FBS, respectively, and cultured at 38.5°C in 5% CO₂/air. When cultures reached 70–80% confluence, cells were disaggregated with 0.5–1.0 mL of Trypsin Replacement Reagent (TrypleSelect^®, Gibco) for further passages, characterization, or cell differentiation.

Cell doubling time in domestic cat AMSCs

AB- and SQ-AMSCs from four domestic cat independent cell cultures were plated in 35-mm tissue culture dishes and expanded from passage 1 (P1) through P5 at an initial concentration of 6000–8000 cells/cm². Cells were passaged at 60–70% confluence for P1 to P4, and at 40–50% confluence for P5. The number of days for a cell to double at each passage was estimated by using the cell doubling time online calculator (www.doubling-time.com/compute.php).

Colony-forming units–fibroblast assay in domestic cat AMSCs

AB- and SQ-AMSCs at P2 were plated in 12-well tissue culture dishes in duplicate serial dilutions at densities of 250, 500, 1000, 2000, 4000, and 8000 cells/cm². Tissue culture medium was changed every 3 days. Colonies that had emerged were counted after Toluidine Blue staining at day 14. Briefly, AB and SQ cells were washed with PBS and stained with 1% Toluidine Blue solution in 2% paraformaldehyde at room temperature for 10 min. Cells were washed several times with water and once with absolute ethanol to remove excess dye. A colony was defined as a distinct blue cluster with round, three-dimensional morphology and no less than 50 cells per colony. The rate of colony-forming units–fibroblast (CFU-F) was calculated by dividing the average number of colonies/well by the total number of cells plated/well.

Expression of surface antigen markers in domestic cat AMSCs

The expression of cluster of differentiation (CD) markers on AB- and SQ-AMSCs was assessed by flow cytometry (FCM) using a protocol reported previously (Gronthos et al., 2001), with minor modifications. Cells at P2–P3 from three cell lines were incubated with primary mouse anti-CD90 (1:50; cat. no. 14-0909, eBioScience, San Diego, CA, USA), anti-CD105 (1:25; cat. no. MCA1557T, AbD Serotec, Raleigh, NC, USA), anti-CD73 (1:25; cat no. 550256, BD Biosciences, San Jose, CA, USA), anti-CD146 (1:25; cat. no. 14-1469, eBioscience, San Diego, CA, USA), anti-CD271 (1:25; cat. no. ab10495, Abcam, Cambridge, MA, USA), anti-CD14 (1:50; cat. no. MCA1568T, AbD Serotec, Raleigh, NC, USA), anti-CD45 (1:50; cat. no. MCA2727T, AbD Serotec, Raleigh, NC, USA), or anti-HLA-DR (1:50; cat. no. 555810, BD Biosciences, San Jose, CA, USA) antibodies.

Antibody-binding reactions were carried out in a blocking solution of 1% bovine serum albumin (BSA) and 5% sheep serum in DPBS at 4°C overnight. Then cells were washed twice in DPBS and incubated with secondary sheep anti-mouse fluorescein isothiocyanate (FITC)-conjugated (1:50; cat. no. S3772, Aldrich, St. Louis, MO, USA) antibody for 2–3 h at room temperature. Finally, cells were washed with 5 mL of DPBS followed by 5 mL of DMEM medium to remove excess/unbound antibodies before FCM analysis. Control for nonspecific binding was performed by incubating cells with sheep immunoglobulin G (IgG) and the secondary conjugated antibody, but not primary antibody. Unstained cells were used as control for autofluorescence.

Cell fluorescence was analyzed with a flow cytometer (FACSAria; Becton Dickinson Immunocytometry Systems, San Jose, CA, USA) using a 488-nm argon-ion laser for excitation and a 530/30-nm bandpass filter for emission collection and logarithmic amplification. Gating parameters were initially set using both unstained and sheep IgG control cells. A minimum of 25,000 events was counted for each sample. All data were analyzed with software Flowjo7.6.5 (Tree Star, Inc., Ashland, OR, USA). Positively stained cells for each CD marker were calculated as gated cells with fluorescence greater than the 99th percentile of the IgG control cells.

Detection of gene transcripts by quantitative RT-PCR in domestic cat AMSCs

The expression of cat pluripotent transcripts Oct-4, Nanog, Klf4, Sox-2, and the internal standard 18SrRNA genes were detected by quantitative RT-PCR (RT-qPCR) performed as described previously (Biancardi et al., 2012; Gómez et al., 2011). Total RNA was isolated from AB- and SQ-AMSCs of four domestic cat cell lines; cat fibroblasts and cat blastocysts served as controls. cDNA was produced by using the RNeasy Kit and QuantiTect Reverse Transcription Kit (Qiagen, Valencia, CA, USA) according to manufacturer's instructions. Briefly, each sample containing about 30,000 cells of either AB- or SQ-AMSCs or cat fibroblasts (control) was lysed, and cell lysates were subjected to an on-column DNase I digestion for 15 min at room temperature to remove genomic DNAs.

Total RNA was eluted from the column with 30–50 μL of RNase-free water and treated with the gDNA Wipeout Buffer to further remove residual genomic DNAs before reverse transcription using a mixture of poly(T)s and random oligomers provided in the QuantiTect Reverse Transcription Kit. cDNA was frozen immediately and stored at −80°C until qRT-PCR. Total mRNA from cat blastocysts was isolated using a Cells-to-cDNA™ II Kit (Ambion, Inc. Austin, TX, USA) according to manufacturer's directions. Briefly, pools of five to six blastocysts were washed in PBS, placed in RNase-free tubes containing 10 μL of cell lysis II buffer, and heated at 75°C for 10 min. Then, 0.06 U/μL of DNase was added to the crude cell lysate and heated at 37°C for 15–20 min to degrade genomic DNA and heated again at 75°C for 5 min to inactivate DNase. Cell lysate was reverse transcribed into cDNA using a one-step RT-qPCR with random oligonucleotides.

The sequences, amplicon sizes, PCR cycles, and accession numbers of the primers used for amplification of the target genes are described in Table 1. RT-qPCR reactions were achieved by the addition of 1.75 μL of cDNA to 40 μL Mastermix containing 30 μL 2 × SYBR Green Supermix (BioRad, Hercules, CA, USA) and 300 nM each of forward and reverse primers, with nuclease-free water. Reactions were performed in single clear tubes in duplicates of 25 μL each, with negative reverse transcription controls to verify absence of genomic contamination and a nontemplate control for each sample and gene, respectively. The cycling conditions were as follows: 95°C for 3 min, followed by 40 cycles of 95°C for 10 sec, 54.2°C for 45 sec, and a final step of 72°C for 5 min using a BioRad iQ5 Multicolor Real-Time PCR system. The ΔΔCt method was used for RT-qPCR data evaluation. Cycle threshold (Ct) values were obtained from the PCR baseline subtracted curve fit analysis, and normalized for different amounts of input cDNA using ΔCt (Ct for the 18SrRNA as the normalizing gene) − (Ct for the gene of interest). Next, ΔΔCt was calculated by subtracting the δCt of each sample from the ΔCt of a reference cDNA sample. The n-fold increase or decrease in expression levels of each gene in AB, SQ-AMSCs, and fibroblast cells was calculated using the formula 2^−ΔΔCt.

Table 1.

Primer Sequences and Conditions for Transcription Detection of Cat Genes

Primer sets	Primer sequences	Amplicon size (bp)	Number of PCR cycles	Accession no. (ref.)
NANOG	F, 5′-cctgaacatacagcctgaagtc-3′	113	29.7	EU366913 (Yu et al., 2009)
	R, 5′-aaggcagagatagaggacccta-3′
KLF4	F, 5′-aagaggaagaaattcaggtacagaaa-3′	111	34.1	HE598712
	R, 5′-agcagtttgagaactttggctt-3′
OCT-4	F, 5′-tgcagctcagtttcaagaaca-3′	112	20–30	EU366914 (Gómez et al., 2008)
	R, 5′-acaagtgtctctgctttgcata-3′
SOX-2	F, 5′-catcacccgcagcaaatga-3′	111	25.5–31.6	HQ385804 (Gómez et al., 2011)
	R, 5′-aagaagtccaggatctctcataaa-3′
18 srRNA	F, 5′- CGGCTACCACATCCAAGGAA-3′	187	26.76	X03205 (Biancardi et al., 2012)
	R, 5′- GCTGGAATTACCGCGGCT-3′

F, forward; R, reverse.

In Vitro multilineage differentiation of black-footed and domestic cat AMSCs

Adipogenic conditions

For adipogenic cell induction, black-footed cat AMSCs (pooled cells isolated from AB and SQ tissue) and domestic cat AB- and SQ-AMSCs at P2–P3 were plated at ∼8000 cells/cm² in six-well tissue culture dishes containing 3 mL of DMEM-F12 medium supplemented with 10% FBS and cultured at 38.5°C in 5% CO₂/air. When cells reached 70–80% confluence, the tissue culture medium was replaced with Adipogenic Induction Medium (AIM) consisting of DMEM-F12 medium supplemented with 10% FBS, 10% rabbit serum, 5 μg/mL insulin, 100 nM dexamethasone, 50 μM arachidonic acid, and 500 μM 3-isobutyl-l-methylxanthine (IBMX), and cultured for 3 additional days. On the fourth day of culture, AIM was switched to Adipogenic Maintenance Medium (AMM), which consists of the same components of the AIM excluding dexamethasone, arachidonic acid, and IBMX. Control cells were cultured in tissue culture medium for the same length of time.

Assessment of adipogenic differentiation was determined semiquantitatively by the accumulation of intracellular oil droplets after Oil Red O staining. On day 5, induced and control cells were fixed with 4% paraformaldehyde for 10 min at room temperature, rinsed once with 70% isopropanol, and stained with 1 mL of 0.6% Oil Red O solution for 15 min. Stained cells were visualized with phase-contrast optics on an inverted microscope (Olympus IX-71). For quantification of staining, the Oil Red O solution in domestic cat AB- and SQ-induced cells was removed and cells were washed with water several times until the washes became clear. Each well was then destained by incubating with 1 mL of 100% isopropanol for 15 min. An empty well was stained as a background staining control. The optical density (OD) of the solution was measured at 520-nm absorbances of the isopropanol extracts on a Spectronic^®Genesys^™ 5 spectrophotometer (Spectronic Instrument, Rochester, NY, USA).

Osteogenic conditions

Black-footed cat AMSCs (pooled cells isolated from AB and SQ tissue) and domestic cat AB- and SQ-AMSCs at P2–P3 were plated at 6000–8000/cm² in a six-well dish containing 3 mL/well of Osteogenic Induction Medium (OIM) consisting of DMEM-F12 medium supplemented with 10% FBS, 0.25 mM ascorbic acid, 100 nM dexamethasone, and 10 mM β-glycerophosphate (Martin et al., 2002). Cells were cultured at 38.5°C in 5% CO₂/air for 21 days, and OIM medium was changed every 2–3 days. The extent of osteogenic differentiation was measured by Alizarin Red S staining of extramatrix calcium deposits. Induced cells were fixed in cold 100% ethanol for 15 min at 4°C. After removing the ethanol and rinsing with water, cells were stained with 1 mL of 40 mM Alizarin Red S (ARS) solution at room temperature for 40 min. Stained cells were visualized with phase-contrast optics on an inverted microscope. For quantification of staining, in domestic cat AB- and SQ-induced cells, the ARS solution was removed and cells were washed with water. A leaching solution consisting of 20% methanol and 10% acetic acid (adjusted to pH 4 with 10% ammonium hydroxide) was used to extract the cells for 30 min at room temperature. An empty well was included for background staining control. The OD of the solution was measured at 450-nm absorbance of the extracts on a Spectronic^®Genesys^™ 5 spectrophotometer.

Chondrogenic conditions

Black-footed cat AMSCs (pooled cells isolated from AB and SQ tissue) and domestic cat AB- and SQ-AMSCs at P2–P3 were plated at ∼8000 cells/cm² in six-well tissue culture dishes containing 3 mL/well of DMEM-F12 medium supplemented with 10% FBS and cultured at 38.5°C in 5% CO₂/air. When cells reached 90–100% confluence, cells were dissociated with 3 mL of TrypleSelect™ (Gibco), and cell numbers counted and centrifuged. Cell pellets were dissociated and ∼500,000 cells transferred into 15-mL polypropylene conical tubes for either chondrogenic induction or negative control treatment, following a protocol previously reported by Estes and Guilak (2011) with minor modifications.

Incomplete Chondrogenic Medium (ICM) for negative control treatment consisted of DMEM and 4.5 g/L glucose medium supplemented with 10% FBS, 10 μL/mL insulin–transferrin–selenium (ITS; Gibco), and 100 nM dexamethasone, whereas Complete Chondrogenic Medium (CCM) for cell induction consisted of the same components as the ICM, but was supplemented with 10 ng/mL recombinant human transforming growth factor-β1 (hTGF-β1; Shenandoah Biotechnology Inc., Warwick, PA, USA) and 1 μM/mL l-ascorbic acid 2-phosphate. Cell mixtures were cultured at 38.5°C in 5% CO₂/air for 4–5 weeks, and tissue culture medium from each treatment was changed every 2–3 days.

Assessment of chondrogenic differentiation was based on Alcian Blue staining of cartilage proteoglycans and by immunocytochemical staining for collagen type II of cartilage-specific extracellular matrix (ECM). Following 4–5 weeks of induction, cell aggregates from each treatment were digested in 300 U of collagenase II and 2 mM CaCl₂ for 90 min, centrifuged at 1500 rpm for 6 min, and plated separately on wells of a tissue culture dish. For Alcian Blue staining (Ullah et al., 2012), differentiated cells attached within 2 h of plating whereas unattached cells (including ECM) were removed. Cells were cultured for 4 additional days and for staining, medium was removed, cells were washed and fixed in cold 4% acetone:methanol solution and stained with 1% Alcian Blue with 3% acetic acid, and then incubated for 30 min followed by three rinses in 3% acetic acid. Stained cells were visualized with an inverted microscope (Olympus IX-71). For quantification of domestic cat AB- and SQ- induced cells, the Alcian Blue staining was solubilized in 1% sodium dodecyl sulfate (SDS) solution for 30 min. The OD of the solution was measured at 605-nm absorbances of the extracts on a Spectronic^®Genesys^™ 5 spectrophotometer.

For collagen type II staining, dissociated differentiated cells from aggregates including some ECM were cultured for 2 additional days before staining. Cells and ECM were then washed and fixed in 4% paraformaldehyde for 20 min at room temperature and incubated with the primary antibody polyclonal rabbit anti-collagen II (1:200; cat. no. AB34712; Abcam, Cambridge, MA USA) at 4°C overnight in a 5% goat serum DPBS blocking solution. Cells were then rinsed and incubated with secondary goat-anti-rabbit Cy3-conjugated antibody (1:200; cat. no. AP187C, Millipore, Temecula, CA, USA) at room temperature for 2–3 h. Control for nonspecific binding was performed by incubating differentiated cells with some ECM with the secondary conjugated antibody, but not the primary antibody. Stained cells were visualized with an inverted microscope (Olympus IX-71).

Neurogenic conditions

Black-footed cat AMSCs (pooled cells isolated from AB and SQ tissue) and domestic cat AB- and SQ-AMSCs at P1–P3 were plated at ∼8000 cells/cm² in a four-well tissue culture dish containing 500 μL/well of DMEM-F12 medium supplemented with 10% FBS and cultured at 38.5°C in 5% CO₂/air. When cells reached 50–60% confluence, tissue culture medium was supplemented with 10 ng/mL human basic fibroblast growth factor (bFGF; Invitrogen, Grand Island, NY, USA) for 24 h prior to induction. DMEM-F12 medium was changed to a Neurogenic Induction Medium (NIM) consisting of DMEM-F12 medium with no serum and supplemented with 2% dimethylsulfoxide (DMSO), 200 μM butylated hydroxyanisole, 25 mM potassium chloride, 2 mM valproic acid, 10 μM forskolin, 1 μM hydrocortisone, 5 μg/mL insulin, and 2 mM l-glutamine (Martin et al., 2002) and cultured at 38.5°C in 5% CO₂/air for 24 h. Control cells were cultured in serum free DMEM-F12 medium for the same length of time.

Assessment of neurogenesis differentiation was based on morphological changes toward neuron-like cells and the expression of the neuronal cell markers anti-microtubule-associated protein 2 (MAP2), anti-β-tubulin III (Tuj-1), and neuronal nuclei protein (NeuN). Induced cells were fixed in 4% paraformaldehyde for 15 min at room temperature, permeablized in 0.2% TritonX-100 PBS solution for 30 min, and incubated with both primary antibodies of polyclonal rabbit anti-human MAP2 (1:500; cat. no. AB5622, Millipore, Temecula, CA, USA) and mouse anti-human NeuN (1:200; cat. no. MAB377, Millipore, Temecula, CA, USA), or with both primary anti-human MAP2 and monoclonal mouse-anti human Tuj-1 (1:1000; cat. no. T8578, Aldrich, St. Louis, MO, USA) at 4°C overnight in a 5% goat serum DPBS blocking solution. Cells were then rinsed with DBPS and incubated for dual staining with secondary goat anti-mouse FITC (1:200, F0257; Aldrich) and goat anti-rabbit Cy3-conjugated (1:200, cat. no. AP187C) antibodies at room temperature for 2–3 h. Also, the extent of neuronal differentiation for each cell type was determined in domestic cat AB- and SQ- induced cells by detecting MAP2 with FCM, which was performed in the same way as described for detection of the expression of CD markers, except that a 575/25-nm bandpass filter was used.

Statistical analysis

A t-test or one-way analysis of variance (ANOVA) was used to analyze the data on the cell doubling time for each cell line and at each passage and the percentage of cells expressing CD markers. The Holm–Sidak method was used to determine differences between two means after ANOVA. Statistical analyses were performed by using SigmaPlot (v. 12.5, Systat Software Inc., San Jose, CA, USA). Differences were considered as significant at p < 0.05.

Results

Isolation, growth, and cell doubling time of cat AMSCs

In the black-footed cat, the amount of adipose tissue collected from SQ tissue was significantly smaller (0.90 g) than that of domestic cats (3.6 ± 0.7 g), possibly due to the thin layer of skin and fat that these cats accumulate under the skin, but the amount of AB biopsy taken was similar (1.0 gram) to that of domestic cats (1.2 ± 0.2 grams). Nonetheless, the numbers of nucleated cells isolated from AB and SQ adipose tissues in the black-footed cat were similar (1.1 × 10⁶ and 1.2 × 10⁶, respectively). Even though in the domestic cat smaller biopsies were obtained from AB adipose tissue, the total mean number of nucleated cells per gram for AB biopsies was variable and similar (0.6–22 × 10⁶) to that of SQ adipose tissue (0.4–24 × 10⁶). Due to the low amount of black-footed cat adipose tissue and the limited availability of black-footed cats to collect further adipose tissue, nucleated cells isolated from AB and SQ adipose tissue were pooled and characterized only for in vitro multilineage cell differentiation.

Within 48–72 h after initial plating, both AB- and SQ-AMSCs attached to the plastic surface and appeared fibroblastic, with a spindle-shaped, polygonal, and elongated surface. Cells from both types of tissue were in culture for 5–8 days before the cells reached 60–70% of confluence and defined as P0. Thereafter, cells required 3–6 days at P1, 6–8 days at P2, 6–11 days at P3, and 9–14 days at P4 to reach 60–70% confluence, and 19–25 days at P5 to reach 50% confluence, when cells started to show signs of senescence. The type of adipose tissue and the genotype did not influence cell doubling time. However, cell doubling time for both types of AMSCs was significantly affected by duration of culture (passage number). In fact, AB- and SQ-AMSCs at P1 (5.02 ± 2.6 vs. 4.47 ± 3.7) and at P2 (7.78 ± 3.7 vs. 6.9 ± 2.0), respectively, required significantly fewer days to complete a cell doubling than that at P3 (10.2 ± 1.1 vs. 10.4 ± 2.9), P4 (13.5 ± 4.8 vs. 15.0 ± 5.2), and P5 (57.6 ± 4.8 vs. 40.1 ± 4.6, respectively, p < 0.05 (Figs. 1 and 2).

FIG. 1.

Scatter plot of days/cell doubling on four populations of domestic cat AMSCs isolated from AB adipose tissue at P1–P5. AB-MSCs at P1 and P2 required significantly fewer days to complete a cell doubling than that at P3, P4, and P5. (**) Significant differences (p < 0.05).

FIG. 2.

Scatter plot of days/cell doubling on four populations of domestic cat AMSCs isolated from SQ adipose tissue at P1–P 5. SQ-MSCs at P1 and P2 required significantly fewer days to complete a cell doubling than that at P3, P4, and all than that at P5. (**) Significant differences (p < 0.05) between P1–P2 and P3–P4; (***) significant differences (p < 0.05) between P1–P2 to P3–P4 and P5.

CFU-F of domestic cat AMSCs

Colonies appeared spontaneously within 12–14 days after initial plating in both types of AMSCs. To confirm that colony formation was not a result of overgrowth, we measured the frequency of AMSCs that were sufficiently proliferative in culture to produce a colony with at least 50 cells in it and were positively stained with Toluidine Blue (Fig. 3). We observed that SQ-AMSCs plated at a density of 1000 cells/cm² formed colonies with the highest frequency rate per number of cells plated (7.1%; Table 2) that that of those plated at higher cell densities. However, no clear differences were observed in colony counts for SQ- and AB-AMSCs plated at cell densities <1000 per cm² (Table 2).

FIG. 3.

Photograph of domestic cat AMSC at 14 days after initial plating and stained with Toluidine Blue. Color images available online at www.liebertpub.com/cell

Table 2.

Colony-Forming Unit–Fibroblast (CFU-F) of Domestic Cat Adipose Mesenchymal Stem Cells (AMSCs) Isolated from Abdominal (AB) and Subcutaneous (SQ) Adipose Tissues and Plated at Serial Dilution Densities

Plating cell density/cm²	SQ-AMSCs CFU-F n (%)	AB-AMSCs CFU-F n (%)
250	—	—
500	—	—
1000	2.5 (7.1)	—
2000	2.5 (3.5)	4 (5.7)
4000	4 (2.8)	4.5 (3.2)
8000	9 (3.2)	4.5 (1.6)

Surface antigen profile of domestic cat AMSCs

Representative histograms of CD markers are shown in Figure 4 and the percentage of positive cells using a panel of eight antibodies on AB- and SQ-AMSCs are summarized in Table 3. AMSCs derived from both tissue types express the typical MSC marker proteins CD90⁺ and CD105⁺. However, CD73, one of the markers expected to be positive, was not detected in either type of AMSCs. AB-AMSCs had higher percentages of cells positive for CD90⁺ (92%) and CD105⁺ (80%) than those of SQ-AMSCs (77%, 58%; respectively, p < 0.05; Table 3). Furthermore, we tested the expression of two antigen surface markers—CD146 to distinguish MSCs from fibroblast cells (Huang et al., 2011), which share a similar morphological appearance, and CD271 to facilitate the detection of MSCs with enhanced colony formation (Bϋhring et al., 2007; Jarocha et al., 2008) and engraftment capacity during stem cell transplantation (Kuçi et al., 2010). FCM analysis revealed a small population of cells positive for CD146⁺ in AB- and SQ-AMSCs (17% and 9%, respectively; p < 0.05); however, CD271⁻ was not detected (Table 3 and Fig. 4). In addition, AMSCs derived from both tissue types displayed no expression of human leukocyte antigen-DR (HLA-DR) and the hematopoietic lineage markers (CD14⁻, CD45⁻), although a small percentage of AB-AMSCs (4%) did show positive expression of CD14⁺ (Table 3).

FIG. 4.

Representative expression of positive and negative markers on four populations of domestic cat cultured AB-AMSCs SQ-AMSCs adipose tissues. AMSCs were stained with monoclonal antibodies directed against either CD90, CD105, CD146, CD73, CD271, CD45, or HLA-DR and coupled with a secondary FITC-conjugated antibody. Histograms show the shift in fluorescence intensities exceeding that of 99% of the IgG control-stained cells (red line) to positively stained samples (blue line). Percentages indicate the percent positive and percent negative cells within the stained sample. Color images available online at www.liebertpub.com/cell

Table 3.

Comparison of the Surface Antigen Profile of Domestic Cat Adipose Mesenchymal Stem Cells (AMSCs) Isolated from Abdominal (AB) and Subcutaneous (SQ) Adipose Tissues and Analyzed by Flow Cytometry

Antigen	AB-AMSCs % positive cells	SQ-AMSCs % positive cells
CD90	92%^a	77%^b
CD105	80%^a	57%^b
CD73	1%	1%
CD146	17%^a	9%^b
CD271	1%	1%
CD14	4%	1%
CD45	2%	1%
HLA-DR	2%	2%
IgG-control	2%	2%

Results are the mean percentage of 20,000 ± 10,000 cells obtained from four donors. Percentage of positive cells is defined as fluorescent intensity greater than 99% of the immunoglobuolin G (IgG) control cells.

Different superscripts within columns indicate significant differences (p < 0.05).

Expression of pluripotent genes in domestic cat AMSCs

To determine whether AMSCs express genes associated with pluripotency, we measured the relative transcript abundance for Oct-4, Nanog, Sox-2, and Klf4 in both AB- and SQ-AMSCs. The RT-qPCR analysis revealed that both cell types expressed Nanog, Klf4, and Oct-4, whereas Sox-2 expression was not detected in either of the AMSCs (Fig. 5). Analysis of pooled data for each cell type showed that the highest expression for Nanog was detected in AB-AMSCs, displaying approximately 15-fold higher expression than control embryos, whereas SQ-ADMSCs showed 13-fold less Nanog than AB-AMSCs and twice that of embryos (Fig. 5). In contrast, expression of KLF4 and Oct-4 by both AB- and SQ-AMSCs was either similar to or lower, respectively, as compared to control embryos (Fig. 5). Also, despite the expected absence of pluripotent stem cell transcripts, RT-qPCR showed detectable levels of Klf4 and Nanog in fibroblast cells (Fig. 5). Expression of pluripotent genes in differentiated cells is not an uncommon situation. In fact, previous studies had reported the expression of Oct-4, Nanog, and KLF4 in human fibroblast cells at similar levels as those present in control cells (Page et al., 2009; Ambadi et al., 2010).

FIG. 5.

Relative transcript abundance of Klf4, Nanog, Oct-4, and Sox-2 genes in cultured domestic cat AMSCs isolated from AB and SQ adipose tissues, in vitro–derived embryos (Em), and cat dermal fibroblasts (Fb). Klf4, Nanog, Oct-4, and Sox-2 genes were normalized with their corresponding 18SrRNA signal, and results are presented as a relative change in ratio among groups. A value of 0 represents no change in transcript abundance. Abundance levels were expressed as an n-fold difference relative to Em which was set as 1. (Jagged arrow) Values are greater than the ones shown in the graph. Error bars depict standard error of the mean (SEM).

Data analysis by individual cell donors (L20, L22, L23, and L33) and tissue depot site from where the cells were isolated (SQ vs. AB; Table 4) clearly showed that the high Nanog levels observed in the analysis of pooled data was directly influenced by one cell culture (L22), whereas both AB- and SQ-AMSCs exhibited 22- to 44-fold higher NANOG transcript expression, respectively, in comparison to that of control cat embryos. In contrast, two cell-cultures (L20 and L32) showed low levels of Nanog transcripts, but only in SQ-AMSCs showing 0.26 and 0.33 times lower Nanog levels than that of cat embryos. A fourth cell culture (L33) did not exhibit Nanog expression in either of the two cell types. Oct-4 was detected (L22) at low levels in only one cell culture of both AB- and SQ-AMSCs, whereas the other cell cultures did not show Oct-4 expression. Interestingly, cell culture L22 expressing Oct-4 was the same cell culture that showed high Nanog transcript expression (Table 4).

Table 4.

Fold Difference in mRNA Expression of Klf4, Nanog, Oct-4, and Sox-2 Genes in Mesenchymal Stromal/Stem Cells (AMSCa) Isolated from Four Cat Cell Donors (Lines 20, 22, 32, 33), from Abdominal (AB) or Subcutaneous (SQ), Relative to Cat Embryos That Were Set as 1

Cell type	Collection site	KLF4	Nanog	Oct-4	Sox-2
Embryo	—	1.00	1.00	1.00	1.00
AMSCs-L20	AB	0.26	0.00	0.00	0.00
	SQ	0.47	0.26	0.00	0.00
AMSCs- L22	AB	N/A	21.89	0.12	0.00
	SQ	N/A	43.94	0.25	0.00
AMSCs- L32	SQ	1.04	0.33	0.00	0.00
AMSCs- L33	AB	1.18	0.00	0.00	0.00
AMSCs- L33	SQ	1.39	0.00	0.00	0.00
Fibroblasts	—	1.12	0.09	0.00	0.00

In vitro multilineage differentiation of black-footed and domestic cat AMSCs

To evaluate the functional capability of AMSCs and possible differences between cells isolated from AB or SQ adipose tissue to give rise to cells of multiple cell lineages, we induced in vitro cell differentiation into mesodermal (osteoblasts, chondrocytes, and adipocytes) and ectodermal (neuronal progenitors) cells by stimulation with inductive factors and compared the extent of differentiation for each cell type.

Mesodermal differentiation

During adipogenic induction, we observed morphological changes into adipose-like cells as early as day 2 of induction. Induced cells became sphere-shaped and the cytoplasm more granular. In the absence of the hormonal cocktail in control cells, no differentiation was observed. Upon fixing and staining with Oil Red O solution, larger intracellular Red Oil droplets were clearly visible, and an equal extent of differentiation was observed between both domestic and black-footed cat AMSCs (Fig. 6). There were no differences between the increases in absorbance (A₅₂₀) of the Oil Red O solution extracts of the induced AB- and SQ-AMSCs (0.46 ± 0.178 vs. 0.45 ± 0.002, respectively). However, the oil deposit inside the induced cells was about 50 × higher as indicated by comparing the A₅₂₀ of the destaining extracts of the induced cells relative to that of the AB- and SQ-AMSC and control cells (0.455 ± 0.01 vs. 0.0 ± 0.0, respectively; p < 0.001).

FIG. 6.

Photomicrographs representing the effect of adipogenic cell induction (treated cells) on domestic cat (DSH) AMSCs isolated from AB and SQ adipose tissues and on black-footed cat (BFC) AMSCs (pooled cells isolated from AB and SQ tissue). Treated cells showed an increased granularity (bright field) and larger intracellular oil droplets stained red by Oil Red. Bright field for BFC treated AMSCs was not included/not available. In the absence of adipogenic induction medium in untreated control cells, no Oil Red O staining was observed. Scale bars, 200 μm. Color images available online at www.liebertpub.com/cell

During osteogenic induction, morphological changes, in both species and cell types of induced AMSCs occurred gradually over 3 weeks. Cell confluence was reached between day 3 and day 5 of induction. After a week of induction, patches of cuboidal cells started to form. By the end of induction at day 21, most of the induced wells were covered with densely packed cuboidal cells and nodules of tightly packed cells (Fig. 7). These nodules tend to have larger calcium phosphate deposits as shown by more intense Alizarin Red S staining. The A₄₅₀ of extracts from destained cells indicated that AB- and SQ-AMSCs (0.065 ± 0.027 vs. 0.076 ± 0.033, respectively) had 1.5 × to 2 × higher calcium deposition in ECM relative to control cells (0.048 ± 0.012 vs. 0.052 ± 0.013, respectively). Similarly, calcium deposition in ECM in black-footed cat AMSCs had 4.2 × higher concentrations (0.156 ± 0.076) relative to control cells (0.048 ± 0.012). Nonetheless, differentiation potentials of domestic AB- and SQ-AMSCs and black-footed cat AMSCs toward osteogenic lineages appear to be similar.

FIG. 7.

Photomicrographs representing the effect of osteogenic cell induction (treated cells) on domestic cat (DSH) AMSCs isolated from AB and SQ adipose tissues and on black-footed cat (BFC) AMSCs (pooled cells isolated from AB and SQ tissue). Treated cells showed large patches of cuboidal cells (bright field) and calcium phosphate stained red by Alizarin Red S. In the absence of osteogenic induction medium in untreated control cells, no cell differentiation (bright field) or Alizarin Red S staining was observed. Scale bars, 200 μm. Color images available online at www.liebertpub.com/cell

During chondrogenic induction, we observed spontaneous formation of cell aggregates by day 2 of induction. The aggregates were dense and spherical in appearance with few outgrowths (Fig. 8). Aggregates derived from the black-footed cat were larger in size than those of domestic AB and SQ, whereas aggregates derived from SQ-AMSCs were slightly larger in size than those derived from AB-AMSCs. However, control uninduced AMSCs also formed cell aggregates of similar size and morphology to those of induced cell aggregates from AB-AMSCs. Aggregates derived from both species and cell types have larger proteoglycan content after 28 days of chondrogenic induction culture as shown by intense Alcian Blue staining (Fig. 8).

FIG. 8.

Photomicrographs representing the effect of chondrocyte cell induction on domestic cat (DSH) AMSCs isolated from AB and SQ adipose tissues and on black-footed cat (BFC) AMSCs (pooled cells isolated from AB and SQ tissue). Aggregates derived from BFC-AMSCs were larger in size than those of DSH-AB and SQ aggregates, whereas aggregates derived from SQ-AMSCs were slightly larger in size than those derived from DSH AB-AMSCs. Aggregates derived from both species and cell types have larger proteoglycan content after 28 days of chondrogenic induction culture as shown by intense Alcian Blue staining. Untreated control AMSCs formed cell aggregates of similar size and morphology to that of induced cell aggregates from AB-AMSCs, but no Alcian Blue staining was observed. Scale bars, 200 μm, except BFC (top), 500 μm. Color images available online at www.liebertpub.com/cell

The A₆₀₅ of extracts from destained cells indicated that induced SQ-AMSCs had higher proteoglycan content in ECM (0.147) than that of AB-AMSCs (0.060), and both were 3.7 × and 1.5 × higher, respectively, relative to control cells (0.039). The differentiation potential of domestic cat SQ-AMSCs toward chondrogenic lineages appears to be higher than that of AB-AMSCs. Immunostaining analysis showed the presence of collagen type II in ECM of chondrogenic-induced aggregates but not on control uninduced aggregates (Fig. 9). Nonetheless, the amount of cartilage-specific ECM in each of the cell types was low possibly due to being lost during dissociation and plating of the cell aggregates.

FIG. 9.

Photomicrographs representing the presence of ECM (collagen type II; red) on dissociated aggregates after 4 weeks of chondrogenic induction or control group without induction. Counterstaining was performed with a DNA fluorescent dye, 4′,6-diamidino-2-phenylindole (DAPI) (blue). DSH, domestic cat; BFC, black-footed cat; BF, bright field. Scale bars, 200 μm. Color images available online at www.liebertpub.com/cell

Ectodermal differentiation

During neurogenic induction, cells from both species (black-footed and domestic cat) and types (AB- and SQ-AMSCs) underwent morphological changes toward neuron-like cells within hours of induction. Induced cells had a compact body with bi- or multipolar, long, thin cytoplasmic extensions that formed intercellular contacts resembling cultured neurons (Fig. 10). After 24 h of induction, few fibroblastic cells remained. Induced neuronal-like cells expressed the pan-neuronal markers MAP2 and Tuj-1 (Fig. 10); however, NeuN expression was not observed. The induction efficiency of domestic AB- and SQ-AMSCs into neuronal-like cells, as measured by positive expression of MAP2 by FCM, indicated that high percentages of the induced AB- (95.3%) and SQ-AMSCs (94.2%) expressed the neuronal marker (Fig. 10).

FIG. 10.

Photomicrographs representing the effect of neurogenic cell induction on domestic cat (DSH) AMSCs isolated from AB and SQ adipose tissues and on black-footed cat (BFC) AMSCs (pooled cells isolated from AB and SQ tissue). BFC and DSH (both cell types; AB and SQ) AMSCs underwent morphological changes toward neuron-like cells within hours of induction (bright field). Induced cells had a compact cell body with bi- or multipolar, long, thin cytoplasmic extensions that formed intercellular contacts resembling cultured neurons. Induced neuronal-like cells expressed the pan-neuronal markers anti-microtubule-associated protein 2 (MAP2; CY3 = red) and anti-β-tubulin III (Tuj-1; FITC = green). DNA stained with DAPI = blue. Induction efficiency of DSH-AB and SQ-AMSCs into neuronal-like cells, as measured by positive expression of MAP2 by FCM (A) indicated that high percentages of the induced AB- (95.3%; blue) and SQ-AMSCs (94.2%; grey) expressed the neuronal marker. IgG control stained cells for AB and SQ-AMSCs showed no fluorescence for MAP2 (B). Scale bars, 200 μm, except lower right, 100μm. Color images available online at www.liebertpub.com/cell

Discussion

Phenotypic and biological characteristics of BM- and adipose-derived MSCs have been reported to be dissimilar; however, it is not clear whether cat AMSCs isolated from different depot sites of adipose tissue are different. Therefore, we compared the biological characteristics of domestic cat AMSCs isolated from AB or SQ adipose tissue, and evaluated the multilineage cell differentiation of black-footed and domestic cat AMSCs for future use in regenerative medicine in domestic and endangered felids.

Proliferation of MSCs is affected by the tissue from which the MSCs are isolated. For instance, domestic cat MSCs isolated from adipose tissue proliferate significantly faster than MSCs isolated from BM (Webb et al., 2012). Conversely, in the present study, we observed that the proliferation rate of domestic cat AMSCs was not affected by the tissue depot site from where the cells were isolated, but was affected by duration in culture. AB- and SQ-AMSCs required a similar number of days to complete a cell doubling at P1 and P2, but required more days to complete a cell doubling at P3–P5, when both types of cells started to show signs of senescence. In addition, it has been shown that domestic cat MSCs isolated from BM do not grow well when they are cultured at low densities, but increasing the initial plating density subsequently improved the frequency of colony formation (Martin et al., 2002).

Similarly, we observed that cell proliferation of AMSCs was reduced at plating of <4000 cells/cm², and colony formation was constant and similar between both cell types at plating densities of >4000 cells/cm². Therefore, to isolate cat AMSCs consistently from each biopsy, and to maintain an adequate proliferation rate during in vitro culture, we plated adipose tissues of ≤1 gram in a 35-mm dish, and tissues of ≥1–3 grams in a 60-mm dish or T25 flask, respectively. Thereafter, we plated a minimum of 8000–10,000 AMSC/cm² at each passage.

Previous studies analyzing phenotypic characteristics of domestic cat MSCs isolated from different tissue types (BM, adipose, and fetal membranes) have indicated that these cells, regardless of the site of collection, express the MSC markers CD90⁺ and CD105⁺ (Estes and Guilak 2011; Martin et al., 2002; Quimby et al., 2011, 2013; Webb et al., 2012; Zhang et al., 2014). However, the expression of CD73, one of the stromal cell markers expected to be positive, has not been detected in domestic cat MSCs (Iacono et al., 2012). Similarly, in the present study, we observed that both domestic cat AB- and SQ-AMSCs contained a high proportion of cells positive for CD90⁺ and CD105⁺, but neither cell type expressed CD73⁻.

Iacono et al. (2012) suggested that the lack of CD73 expression by cat MSCs isolated from fetal fluid/membranes and cat circulating lymphocytes may be explained by the inability of the human antibody source to cross-react with the corresponding cat epitopes. Likewise, the commercial antibody used in our study was produced against human CD73, so the negative reactivity of cat MSCs to the antibody for CD73 may be due partially to the type of antigen epitope or isoforms recognized by the antibody, which may be different between human and cat. However, measuring the relative expression of CD73 at the gene level, Rutigliano et al. (2013) reported that multipotent progenitor-like amniotic epithelial cells isolated from cat amnion tissue did not express the CD73 gene (5′-nucleotidase, ecto) but did express other MSC genes. Thus, the negative expression of CD73 in our domestic cat AMSCs may be due to species-specific differences in cell-surface markers expression, which is dynamic and heterogeneous (Pierantozzi et al., 2011).

Several surface markers have been used to characterize subpopulations of MSCs. CD146 is a surface marker important for endothelial cell migration and angiogenesis (Zeng et al., 2012), and it is expressed by human pericyte cells (proposed to be precursors of AMSCs; Crisan et al., 2008; Zimmerlin et al., 2013), a population of cells in a “transitional stage” between pericytes and undifferentiated AMSCs (Zimmerlin et al., 2013), and MSCs isolated from several human tissues (Covas et al., 2008); however, a low number of fibroblast cells express the CD146 marker (Covas et al., 2008; Halfon et al., 2011). Therefore, due to the low number of fibroblast cells positive for CD146⁺, it was suggested that CD146 may be a good marker to distinguish MSCs from fibroblast cells (Huang et al., 2011). To characterize the phenotype of domestic cat AMSCs further and possibly to isolate these cells from fibroblast cells, we evaluated by FCM the expression of CD146 in domestic cat AMSCs isolated from AB and SQ adipose tissue. Analysis revealed that small populations of domestic cat AB- and SQ-AMSCs (17% and 9%, respectively) were positive for CD146⁺. Even though we observed a subpopulation of cat AMSCs positive for CD146⁺, we cannot conclude that this subpopulation of cells specifically is AMSCs.

It is known that the Ad-SVF contains three major populations of stem/progenitors cells—endothelial progenitor cells, pericytes, and AMSCs, as well as fibroblast cells within other cell types (Bourin et al., 2013). In a recent study, Zimmerlin et al. (2013) characterized the expression of stem/progenitor subpopulations in the human Ad-SFV. FCM analysis indicated that a fraction of undifferentiated human AMSCs positive for CD34⁺ are negative for CD146⁻ and the endothelial marker CD31⁻, whereas the phenotypes for pericyte cells (CD34⁻/CD146⁺/CD31⁻) and transitional cells (CD34⁺/CD146⁺/CD31⁻) differ from undifferentiated AMSCs and express the CD146⁺ marker. However, all three cell types co-express MSC-specific markers. Although, in the present study, we did not test simultaneously for the co-expression of several surface markers, we suggest that domestic cat AB- and SQ-AMSCs positive for CD146⁺ and MSCs markers CD90⁺/CD105⁺ may be a heterogeneous subpopulation possibly containing a subset of progenitor cells of pericyte origin or at various stages, and perhaps for AMSCs, as Covas et al. (2008) indicated, that human MSCs express CD146⁺. Moreover, the positive expression of CD146⁺ in cat fibroblast cells (data not shown) suggests that CD146 may not be an ideal surface marker for isolating fibroblast cells from stem/progenitor subpopulations in the cat Ad-SFV.

We also tested domestic cat AMSCs for the expression of CD271, another cell-surface marker that has been used in humans for identifying MSCs with high proliferative rates, clonogenic and enhanced colony formation capacities, as well as potential ability to differentiate into other cell lineages (Huang et al., 2011). Analysis showed that domestic cat AMSCs do not express CD271⁻ for reasons that are similar to those that may explain the absence of CD73. The lack of specific markers in the cat, and differences in expression of MSCs markers between species, increase the difficulty of characterizing cat AMSCs. So far, the consensus is that cultured cat AMSCs are characterized as CD90⁺/CD105⁺/CD45⁻/CD14⁻ with a subpopulation of cells expressing CD146⁺.

In addition to characterizing the general phenotypic characteristics of domestic cat AMSCs as described above, we also evaluated the ability of domestic and black-footed cat AMSCs to differentiate into other cell types. Several studies have reported that lipogenesis activity and gene expression vary between depot sites of adipose tissue (Garaulet et al., 2006; Maslowska et al., 1993; Rodriguez et al., 2002), and that these characteristics may further affect the developmental abilities of MSCs isolated from different adipose tissue depots. For example, in rabbits, the osteogenic potential of AMSCs isolated from the visceral fat of AB was reported to be greater than that of AMSC isolated from SQ adipose tissue (Peptan et al., 2006). In contrast, in humans, osteogenesis is more robust in SQ-AMSCs from the flank and thigh, as compared with that of AMSCs from AB adipose tissue (Levi et al., 2010).

In the present study, we did not observe that the adipose site depot from which AMSCs were isolated influenced the extent of differentiation, with the exemption of AMSCs isolated from SQ adipose tissue that appear to have better differentiation potential toward chondrogenic lineages than that of AMSCs isolated from AB adipose tissue. Nonetheless, the low sample numbers (n = 4) tested and the better capability of black-footed cat AMSCs to differentiate also toward chondrocytes (where AMSCs were pooled from both adipose tissue depots) suggest that differences among individuals account more for the capability to differentiate toward specific cell types than that of the adipose tissue depot from which AMSCs were isolated. Under appropriate stimuli, AMSCs from both the domestic and black-footed cats had greater and more robust adipogenesis ability than they did toward osteogenesis. This concurs with reports from other mammalian species in which it was observed that AMSCs tend to be less “responsive” toward osteogenesis than adipogenesis (Zachar et al., 2011).

It is not clear why domestic cat and black-footed cat AMSCs have better ability to differentiate toward adipocytes, but a previous study in humans indicated that AMSCs have greater ability to differentiate toward adipogenesis than toward osteogenesis or chondrogenesis (Liu et al., 2006). The authors compared the transcriptome profile of human AMCSs vs. BM-MSCs and showed that the expression of genes related to energy reserve and cholesterol metabolism are upregulated in AMSCs, whereas the expression of genes associated with biomineralization and osteoblastic differentiation are significantly lower in comparison to that of BM-MSCs. They concluded that the greater ability of AMSCs to differentiate into adipocytes may be due to the fact that AMSC cultures are dominated by adipogenic progenitors. Therefore, we suggest that the robust ability of cat AMSCs toward adipogenesis may be due in part to the presence of adipogenic cell progenitors and, although we did not evaluate the expression of adipogenic genes in our cell cultures, higher expression of adipogenic transcription factors have been reported previously in domestic cat AMSCs (Zhang et al., 2014).

In addition to differentiation toward mesodermal cell types, domestic cat and black-footed cat AMSCs differentiated into putative neurogenic cells, exhibiting a neuronal-like morphology and expressing several proteins consistent with the neuronal phenotype, as previously described for domestic cat BM-MSCs (Martin et al., 2002). Most of the domestic and black-footed cat neuronal induced cells expressed the neuronal Map-2 and Tuj-1 markers. However, NeuN expression was not observed, a marker that appears to be expressed in the neuron at withdrawal from the cell cycle and/or with the initiation of terminal differentiation. We have demonstrated that high numbers of cat ESCs differentiate into neuronal phenotype cells, with the majority of the neuronal induced cells expressing Map-2, and a low number of induced cat ESCs expressing the neuronal NeuN marker (Gómez and Pope, 2014). It is possible that the differences in expression of NeuN between induced cat ESCs and induced cat AMSCs may be because transdifferentiation into ectoderm lineages by MSCs is expected to be a more challenging process (Zuk et al., 2001). The expression of additional neuronal genes by immunofluorescence or RT-PCR is necessary to characterize the induced neuron-like cells further from cat AMSCs.

The lower ability to differentiate toward neuronal cells may be due in part to the limited multipotent ability of AMSCs. We evaluated the expression of pluripotent markers Nanog, Oct-4, Sox-2, and Klf4 in AMSCs isolated from AB and SQ adipose tissue. RT-qPCR analysis showed that the transcription factor Oct-4 was detected only in one of the cell cultures, at lower levels to that of control embryos, but Sox-2 was not detected in any of the cell cultures. In contrast, the transcript levels of Nanog varied between cell donors. Low levels of Nanog transcripts were detected in two cell cultures and high transcript levels were detected in one cell culture as compared to that of control embryos. A recent study evaluating the expression of pluripotent markers Nanog and Oct-4 in one set of domestic cat progenitor-like amniotic epithelial cells showed that Oct-4 was expressed by these cells at all passages analyzed (P2–P9), whereas only low levels of Nanog expression were detected at P2, P3, and P7 (Rutigliano et al., 2013).

Although it is not possible to compare Rutigliano et al. (2013) results with ours due to differences in cell types, method of culture, and tissue culture media, it is clear that the expression of Oct-4 and Nanog in both studies could be due, in addition to the above-mentioned factors, to the genetic variability between donors from which AMSCs were isolated (Baer et al., 2013). The overexpression of Nanog observed in one of the cell cultures is particularly interesting because it has been reported to be an essential master transcription factor for the reprogramming of differentiated cells toward induced pluripotent stem cells in multiple feline species (Verma et al., 2013). Recently, it was reported that the composition of the medium for culturing AMSCs induced upregulation of the pluripotent genes Nanog, Oct-4, Sox-2, and Rex-1 in human cultured AMSCs (Baer et al., 2010). The biological significance of Nanog overexpression in one of the cat AMSCs cell culture is not clear, but it is possible that the culture medium and or culture methods are influencing the over-expression of the pluripotent marker (Ambadi et al., 2010).

Overall, our results indicated that domestic cat and black-footed cat AMSCs can be isolated from the Ad-SFV and, regardless of the site of adipose tissue depot, have the ability to differentiate into other cell types of ectodermal and endodermal lineages. However, reprogramming efficiencies were lower for differentiation to certain cell types, perhaps an indication of the heterogeneity of the AMSCs population, the inability of cat AMSCs to differentiate fully into mature cells of another cell type, or the induction protocols used in our study were not the most appropriate for cats. Understanding the signaling pathways required to induce differentiation of cat AMSCs into mature functional cell types may facilitate the use of differentiated cells at different stages of maturity as therapeutic agents for diseased or injured domestic or endangered felids.

Footnotes

Acknowledgments

We thank Dr. Michal Soosaar, Dr. Elsburgh Clarke, and Amanda Franklin for performing the surgical procedures and to animal care personnel at the Audubon Center for Research of Endangered Species (ACRES) for the care and husbandry of domestic cats. The assistance of Constance Porretta with flow cytometry at the Louisiana State University Health Sciences Center is greatly appreciated. This study was funded in part by a grant from ACRES and Louisiana State University System collaborative projects.

Author Disclosure Statement

The authors declare that no conflicting financial interests exist.

References

Ambadi

, Malcuit

, Kashpur

, Kole

, Holmes

W.F.

, Hedblom

, Page

R.L.

, and Dominko

(2010). Expression of NANOG and NANOGP8 in a variety of undifferentiated and differentiated human cells. Int. J. Dev. Biol., 54, 1743–1754.

Aust

, Devlin

, Foster

S.J.

, Halvorsen

Y.D.

, Hicok

, Du

Laney

, T., Sen

, Willingmyre

G.D.

, and Thomas

E.D.

(2004). Yield of human adiposederived adult stem cells from liposuction aspirates. Cytotherapy, 6, 7–14.

Baer

P.C.

, Griesche

, Luttmann

, Schubert

, Luttmann

, and Geiger

(2010). Human adipose-derived mesenchymal stem cells in vitro: Evaluation of an optimal expansion medium preserving stemness. Cytotherapy, 12, 96–106.

Baer

P.C.

, Kuci

, Krause

, Kuci

, Zielen

, Geiger

, Bader

, and Schubert

(2013). Comprehensive phenotypic characterization of human adipose-derived stromal/ stem cells and their subsets by a high throughput technology. Stem Cells Dev., 22, 330–339.

Biancardi

M.N.

, Pope

C.E.

, Powell

R.H.

, Galiguis

J.L.

, Dumas

, Morris

, and Gómez

M.C.

(2012). Evaluation of reference gene transcript stability during early pre-implantation development and in spermatogonial cells in the cat. Repro. Fertil. Dev., 24, 187–188.

Bourin

, Bunnell

B.A.

, Casteilla

, Dominici

, Katz

A.J.

, March

K.L.

, Redl

, Rubin

J.P.

, Yoshimura

, and Gimble

J.M.

(2013). Stromal cells from the adipose tissue-derived stromal vascular fraction and culture expanded adipose tissue-derived stromal/stem cells: A joint statement of the International Federation for Adipose Therapeutics and Science (IFATS) and the International Society for Cellular Therapy (ISCT). Cytotherapy, 15, 641–648.

Bϋhring

H.J.

, Battula

V.L.

, Treml

, Schewe

, Kanz

, and Vogel

(2007). Novel markers for the prospective isolation of human MSC. Ann. NY Acad. Sci., 1106, 262–271.

Covas

D.T.

, Panepucci

R.A.

, Fontes

A.M.

, Wilson

A.S.

, Orellana

M.D.

, Freitas

M.C.C.

, Neder

, Santos

A.R.D.

, Peres

L.C.

, Jamur

M.C.

, and Zago

M.A.

. (2008). Multipotent mesenchymal stromal cells obtained from diverse human tissues share functional properties and gene expression profile with CD146+ perivascular cells and fibroblasts. Exp. Hematol., 36, 642–654.

Crisan

, Yap

, Casteilla

, Chen

C.W.

, Corselli

, Park

T.S.

, Andriolo

, Sun

, Zheng

, Zhang

, Norotte

, Teng

P.N.

, Traas

, Schugar

, Deasy

B.M.

, Badylak

, Buhring

H.J.

, Giacobino

J.P.

, Lazzari

, Huard

, and Peault

(2008). A perivascular origin for mesenchymal stem cells in multiple human organs. Cell Stem Cell, 3, 301–313.

10.

Driscoll

C.A.

, Menotti-Raymond

, Roca

A.L.

, Hipe

, Johnson

W.E.

, Geffen

, Harley

E.H.

, Delibes

, Pontier

, Kitchener

A.C.

, Yamaguchi

, O'Brien

S.J.

, and Macdonald

D.W.

(2007). The near eastern origin of cat domestication. Science, 317, 519–523.

11.

Estes

B.T.

, and Guilak

(2011). Three-dimensional culture systems to induce chondrogenesis of adipose-derived stem cells. In: Adipose-Derived Stem Cells: Methods and Protocols, Methods in Molecular Biology. Gimble

J.M.

and Bunnell

B.A.

, eds. (Springer Science-Business Media) pp. 201–206.

12.

Fortier

L.S.

, and Travis

A.J.

(2011). Stem cells in veterinary medicine. Stem Cell Res. Ther. 2, 9.

13.

Garaulet

, Hernandez-Morante

J.J.

, Lujan

, Tebar

F.T.

, and Zamora

(2006). Relationship between fat cell size and number and fatty acid composition in adipose tissue from different fat depots in overweight/obese humans. Int. J. Obes., 30, 899–905.

14.

Gimble

J.M.

, Katz

A.J.

, and Bunnell

B.A.

(2007). Adipose derived stem cells for regenerative medicine. Circ. Res., 100, 1249–1260.

15.

Gómez

M.C.

, Pope

C.E.

, Kutner

R.H.

, Ricks

D.M.

, Lyons

L.A.

, Ruhe

, Dumas

, Lyons

, Lopez

, Dresser

B.L.

, and Reiser

(2008). Nuclear transfer of sand cat cells into enucleated domestic cat oocytes is affected by cryopreservation of donor cells. Cloning Stem Cells, 10, 469–484.

16.

Gómez

M.C.

, Pope

C.E.

, Biancardi

M.N.

, Dumas

, Galiguis

, Morris

A.C.

, Wang

, and Dresser

B.L.

(2011). Trichostatin A modified histone covalent pattern and enhanced expression of pluripotent genes in interspecies black-footed cat cloned embryos but did not improve in vitro and in vivo viability. Cell. Reprogram., 13, 315–329.

17.

Gómez

M.C.

, and Pope

C.E.

(2014). Pluripotent and multipotent domestic cat stem cells: Current knowledge and future prospect. In: Stem Cell Biology and Regenerative Medicine; Stem Cells in Animal Species: From Pre-Clinical to Biodiversity. Brevini

, ed. (Springer Cham, New York, NY) pp. 119–142.

18.

Gronthos

, Franklin

D.M.

, Leddy

H.A.

, Robey

P.G.

, Storms

R.W.

, and Gimble

J.M.

(2001). Surface protein characterization of human adipose tissue-derived stromal cells. J. Cell. Physiol., 189, 54–63.

19.

Halfon

, Abramov

, Grinblat

, and Ginis

(2011). Markers distinguishing mesenchymal stem cells from fibroblasts are downregulated with passaging. Stem Cells Dev., 20, 53–66.

20.

Huang

, Leung

, Peng

, Li

, Lu

F.J.

, Wang

, Lu

, Cheung

K.M.C.

, and Zho

(2011). Developmental definition of MSCs: New insights into pending questions. Cell. Reprogram., 13, 465–472.

21.

Iacono

, Cunto

, Zambelli

, Ricci

, Tazzari

P.L.

, and Merlo

(2012). Could fetal fluid and membranes be an alternative source for mesenchymal stem cells (MSCs) in the feline species? A preliminary study. Vet. Res. Commun., 36, 107–118.

22.

Jarocha

, Lukasiewiez

, and Majka

(2008). Advantage of mesenchymal stem cells (MSC) expansion directly from purified bone marrow CD105+ and CD271+ cells. Folia Histochem. Cytobiol., 46, 307–314.

23.

Kuçi

, Kuçi

, Kreyenberg

, Deak

, Pütsch

, Huenecke

, Amara

, Koller

, Rettinger

, Grez

, Koehl

, Latifi-Pupovci

, Henschler

, Tonn

, Von

Laer

, D., Klingebiel

, and Bader

(2010). CD271 antigen defines a subset of multipotent stromal cells with immunosuppressive and lymphohematopoietic engraftment-promoting properties. Haematologica, 95, 651–659.

24.

Levi

, James

A.W.

, Glotzbach

J.P.

, Wan

D.C.

, Commons

G.W.

, and Longaker

M.T.

(2010). Depo-specific variation in the osteogenic and adipogenic potential of human adipose-derived stromal cells. Plast. Reconstr. Surg., 126, 822–834.

25.

Liu

T.M.

, Martina

, Hutmacher

D.W.

, Hui

J.h.

, Lee

E.H.

, and Lim

(2006). Identification of common pathways mediating differentiation of bone marrow- and adipose tissue-derived human mesenchymal stem cells into three mesenchymal lineages. Stem Cells, 25, 879–886.

26.

Martin

D.R.

, Cox

N.R.

, Hathcock

T.l.

, Niemeyer

G.P.

, and Baker

H.J.

(2002). Isolation and characterization of multipotential mesenchymal stem cells from feline bone marrow. Exp. Hematol., 30, 879–886.

27.

Maslowska

M.H.

, Sniderman

A.D.

, MacLean

L.D.

, and Cianflone

(1993). Regional differences in triacylglycerol synthesis in adipose tissue and in cultured preadipocytes. J. Lipid Res., 34, 219–228.

28.

Page

R.L.

, Ambady

, Holmes

W.F.

, Vilner

, Kole

, Kashpur

, Huntress

, Vojtic

, Whitton

, and Dominko

(2009). Induction of stem cell gene expression in adult human fibroblasts without transgenes. Cloning Stem Cells, 11, 417–426.

29.

Peptan

I.A.

, Hong

, and Mao

J.J.

(2006). Comparison of osteogenic potentials of visceral and subcutaneous adipose-derived cells of rabbits. Plast. Reconstr. Surg., 117, 1462–1470.

30.

Pierantozzi

, Gava

, Manini

, Roviello

, Marotta

, Chiavarelli

, and Sorrentino

(2011). Pluripotency regulators in human mesenchymal stem cells: Expression of NANOG but not of OCT-4 and SOX-2. Stem Cells Dev., 20, 915–923.

31.

Pope

C.E.

(2004). In vitro fertilization and embryo transfer in felids. In: Methods in Molecular Biology, vol 254: Germ Cell Protocols, Vol 2: Molecular Embryo Analysis, Live Imaging, Transgenesis and Cloning. Schatten

, ed. (Humana Press Inc., Totowa, New Jersey) pp. 227–244.

32.

Pope

C.E.

, Gómez

M.C.

, and Dresser

B.L.

(2006a). In vitro embryo production and embryo transfer in domestic and non-domestic cats. Theriogenology, 66, 1518–1524.

33.

Pope

C.E.

, Gómez

M.C.

, and Dresser

B.L.

(2006b). In vitro production and transfer of cat embryos in the 21^st century. Theriogenology, 66, 59–71.

34.

Quimby

J.M.

, Webb

T.L.

, Gibbons

D.S.

, and Dow

S.W.

(2011). Evaluation of intrarenal mesenchymal stem cell injection for treatment of chronic kidney disease in cats: A pilot study. J. Feline Med. Surg., 13, 418–426.

35.

Quimby

J.M.

, Webb, T.l., Habenicht

L.M.

, and Dow

S.W.

(2013). Safety and efficacy of intravenous infusion of allogeneic cryopreserved mesenchymal stem cells for treatment of chronic kidney disease in cats: results of three sequential pilot studies. Stem Cell Res.Ther. 4, 48.

36.

Rodriguez

V.M.

, Pico

, Portillo

M.P.

, Macarulla

M.T.

, and Palou

(2002). Fat source regulates ob gene expression in white adipose tissue of rats under hyperphagic feeding. Br. J. Nutr., 87, 427–434.

37.

Ribitsch

, Burk

, Delling

, Geißler

, Gittel

, Jülke

, and Brehm

(2010). Basic science and clinical applications of stem cells in veterinary medicine. Adv. Biochem. Engin. Biotechnol., 123, 219–263.

38.

Rutigliano

, Corradetti

, Valentini

, Bizarro

, Meucci

, Cremonesi

, and Lange-Consiglio

(2013). Molecular characterization and in vitro differentiation of feline progenitor-like amniotic epithelial cells. Stem Cell Res.Ther. 4, 133.

39.

Terio

K.A.

, O'Brien

, Lamberski

, Famula

T.R.

, and Munson

(2008). Amyloidosis in black-footed cats (Felis nigripes). Vet. Pathol., 45, 393–400.

40.

Trzil

J.E.

, Masseau

, Webb

T.L.

, Chang

C.H.

, Dodam

J.R.

, Cohn

L.A.

, Liu

, Quimby

J.M.

, Dow

S.W.

, and Reinero

C.R.

(2014). Long term evaluation of mesenchymal stem cell therapy in a feline model of chronic allergic asthma. Clin. Exp. Allergy., 44, 1546–1557.

41.

Ullah

, Hamouda

, Stich

, Sittinger

, and Ringe

(2012). A reliable protocol for the isolation of viable, chondrogenically differentiated human mesenchymal stem cells from high-density pellet cultures. BioRes. Open Access, 1, 297–305.

42.

Verma

, Liu

, Holland

M.K.

, Temple-Smith

, Williamson

, and Verma

P.J.

(2013). Nanog is an essential factor for induction of pluripotency in somatic cells from endangered felids. BioResearch, 2, 72–76.

43.

Volk

S.W.

, and Theoret

(2013). Translating stem cell therapies: The role of companion animals in regenerative medicine. Wound Repair Regen., 21, 382–394.

44.

Webb

T.L.

, Quimby

J.M.

, and Dow

S.W.

(2012). In vitro comparison of feline bone marrow-derived and adipose tissue-derived mesenchymal stem cells. J. Feline Med. Surg., 14, 165–168.

45.

Wongsrikeao

, Saenz

, Rinkoski

, Otoi

, and Poeschla

(2011). Antiviral restriction factor transgenesis in the domestic cat. Nat. Methods, 8, 853–859.

46.

X.F.

, Kin

J.H.

, Jung

E.J.

, Jeon

J.T.

, and Kong

I.K.

(2009). Cloning and characterization of cat POU5F1 and Nanog for identification of embryonic stem-like cells. J. Reprod. Dev., 55, 361–366.

47.

Zachar

, Rasmussen

J.G.

, and Fink

(2011). Isolation and growth of adipose tissue-derived stem cells. Methods Mol. Biol., 698, 37–49.

48.

Zeng

, Li

, Lu

, Wu

, Duan

, Luo

, Feng

, Yang

, Fu

, and Yan

(2012). CD146, an epithelial-mesenchymal transition inducer, is associated with triple-negative breast cancer. Proc. Natl. Acad. Sci. USA, 109, 1127–1132.

49.

Zhang

, Dietrich

M.A.

, and Lopez

M.J.

(2014). Therapeutic doses of multipotent stromal cells from minimal adipose tissue. Stem Cell Rev. Rep., 10, 600–611.

50.

Zimmerlin

, Donnenberg

V.S.

, Rubin

J.P.

, and Donnenberg

A.D.

(2013). Mesenchymal markers on human adipose stem/progenitor cells. Cytometry Part A, 83A, 134–140.

51.

Zuk

P.A.

, Zhu

, Mizuno

, Huang

, Futrell

J.W.

, Katz

A.J.

, Benhaim

, Lorenz

H.P.

, and Hedrick

M.H.

(2001). Multilineage cells from human adipose tissue: Implications for cell-based therapies. Tissue Eng., 7, 211–228.